The cerebral cavernous malformation (CCM) is a common hemorrhagic vascular anomaly, presenting in sporadic and familial autosomal dominant forms. It affects more than a million Americans, predisposing them to a lifetime risk of stroke and epilepsy. There is currently no therapy to prevent the genesis or clinical progression of CCM lesions. We have developed novel models of CCM in mice heterozygous for two of the CCM familial genes (Ccm1 and 2), and these exhibit CCM lesions recapitulating the human disease. Our group and others have implicated RhoA/ROCK signaling in vascular hyperpermeability in heterozygous Ccm1 and Ccm2 mice. This hyperpermeability can be reversed by the ROCK inhibitor, fasudil, or by simvastatin which also impacts this same pathway. And we provided evidence of increased ROCK activity in endothelial cells (EC) of murine lesions, and in surgically resected human CCM lesions. We hypothesize that aberrant RhoA activation is the ultimate cause of CCM pathogenesis, such that in vivo ROCK inhibition will inhibit lesion development and hemorrhage. We propose to compare the effect of selective ROCK inhibition with fasudil, versus broader Rho inhibition with simvastatin, given mechanistic differences and potential clinical advantages of each drug. We further hypothesize that ROCK inhibition impacts molecular biomarkers and related gene expression in relevant CCM vascular targets, validating the therapeutic effect. Three aims are proposed to test these hypotheses. In Specific Aim 1, we examine the effect of fasudil or simvastatin at high and low doses on CCM lesion burden, hemorrhage and phenotype, with onset of treatment early and later in life, and we compare therapeutic effect in murine models CCM1 and 2. In Specific Aim 2, we examine the effect of high and low doses of fasudil or simvastatin on established murine CCM lesions pre-screened by in vivo MRI, including effects on lesion size, hemorrhage, phenotypic markers, and the differential expression of genes related to ROCK and associated signaling pathways. And in Specific Aim 3 we examine biomarkers of ROCK activation in surgically resected human CCM lesions, validated by correlative gene expression, as a potential target of therapy in man. We propose to compare lesions from sporadic and familial backgrounds, and with each of three known human gene loci. This translational strategy leverages a number of mechanistic discoveries and recently optimized models, to develop a novel and viable therapy for CCM.